To balance the scales of justice, Lady Justice must be blindfolded to reduce bias. In enterprise computing, capacity planning analysts can't balance system resources to achieve optimal performance and efficiency with a blindfold on. In this fourth part of our series on mainframe-related IT roles, Chuck Hopf, capacity planning analyst with Merrill Consultants, shared his take on his role and how it is similar to that of a database administrator (DBA).
Who Are Capacity Planning Analysts and What Do They Do?
The management discipline of capacity planning ensures that sufficient resources are in place to satisfy the organization's computing workload. Like DBAs, capacity planners need to be detailed and results-oriented and must be able to communicate clearly with all stakeholders. They also must have strong reporting and statistical skills to interpret and use complex data and industry trends to influence decision-making processes related to computing resources. Capacity planning requires a knowledge of software and hardware requirements, learning how to balance system needs with customer and business demands, and ensuring projects are completed in a timely manner with minimal down time. This also requires an ability to coordinate with executives, project leaders, and other stakeholders on a project's metrics and computing resource needs.
"Part of the job," says Hopf, "is just knowing how much capacity you need to do something." He adds that capacity planners monitor a lot of enterprise-wide data all at once, and often with multiple monitors going at the same time. Accurate data is essential to this job because without it, it is difficult to identify cost and resource-optimization opportunities. Data is necessary to understand how to optimize existing resources to ensure applications remain online and that all business operations are adequately supported.
In the past, planners had to rely on multiple tools to collect, analyze, and report data, but IBM's Performance and Capacity Analytics can provide data from applications and workloads, even those in hybrid clouds, enabling capacity planners to forecast future capacity needs when businesses adopt new programs or hardware or when demand patterns change.
Real-time data leveraged from structured logs across the enterprise enables planners to garner a more holistic view of performance and improve critical decision-making. Coupled with industry trends and other information, planners can then leverage the data to align utilization and capacity with expected growth and adapt more easily to future resource needs.
Capacity Planning Analyst Skills Requirements
In most cases, capacity planners have a bachelor's degree in economics, mathematics, or another quantitative field of study. A computer science degree can be beneficial as it provides a potential employee with knowledge of operating systems and hardware. Hopf explained that "on-the-job training offers employees the hands-on knowledge they need to effectively forecast capacity and resource decisions."
Planners may want to consider courses from outside providers, such as ProTech and Interskill, or educating themselves with IBM's product manuals so they are familiar with workload and performance management, as well as forecasting models and data analytics.
Hopf adds that knowing each silo’s language is important. For instance, executives are focused on cost, reputation, customers, and revenue, and facilities staff are focused on power, space, and cooling needs. Application developers, on the other hand, often discuss performance, response time, and transaction volumes, while other technologists are looking at utilization, hardware, and availability.
Capacity Planning Requires Collaboration and Testing
In order to gauge capacity and performance, systems need to be tested for duration of time in processing data. With the multitude of transactions businesses perform daily, it can be hard to test all data loads, which is why a subset should be tested for planners to extrapolate a system's performance level and whether resources need to be shifted to other applications or programs. In a world of shifting priorities and needs, flexibility is key.
"It's more like art than science," says Hopf, "especially when enterprises want to deploy new applications quickly." He adds that this "need for speed" leaves capacity planners with little time to test before applications go live.
According to Hopf, "We have to know a little bit about everything going on in the mainframe, including information about the databases being used and how operations work." Having this knowledge and the language used by other units can make communication easier, especially when operational systems are upgraded, applications are added, or other changes are forthcoming. He adds that it is very important to review the previous day's jobs and data to determine at what capacity systems are running and if adjustments are necessary. "When developers, for instance, communicate changes to planners, capacity issues can be mitigated, either through changes in other programs or applications or through new capacity investments, which could include the purchase of additional hardware," says Hopf.
Balance the Scales With Data
Balancing the resource and capacity scales at an enterprise is a Herculean task, especially given the amount of data mainframes and other systems are producing and using on a daily basis. "Capacity planning is all about the data; the more accurate, the better," Hopf says. He advises that new capacity planners understand the ins-and-outs of their enterprise-wide operating systems, make connections with database administrators and developers so that changes coming down the pike are caught early on in the process, and learn continuously about the latest software in development.